9 THE DEVELOPING VISUAL PATHWAYS Human vision requires precise collaboration of diverse structures. From the eye to the cerebral cortex, the components mature in parallel, each influencing the development of the whole. Some developmental processes follow an innate plan that is programmed using molecular cues forming “hard-wired” neural circuits. Others are controlled by the neuronal activity within the system itself that arises spontaneously, or from visual stimulation. Thus, the anatomical configuration of the visual system is sculpted by both nature and nurture. In a developing individual, visual experience adjusts the neural structures such that they best represent the world they are exposed to. Combined with the innate, “hard-wired,” plan, this produces an efficient visual system because only elements that function appropriately are maintained: “use it or loose it.” Reliance on visual experience makes the system vulnerable: a fault during development may be detrimental. With anomalous visual experience, the system develops abnormally. Thus, the processes that normally generate an efficient visual system can cause abnormal development. An example is monocular deprivation. Here, a problem that obscures vision in one eye, like a congenital cataract, causes abnormal visual development. The visual system allots less cortical tissue to the deprived eye, generating a permanent visual loss that persists after removal of the cataract. The interdependent elements of the visual system must all develop appropriately. A fault at any point from eye to brain can have effects on the whole system. The normal development of each component will be discussed individually. The eye The eyes differentiate early from the neural plate. They first appear as the optic pits by the fifth week of gestation, and then extend from the neural tube to form the optic vesicles. These spherical pouches invaginate to form the optic cups, attached to the pros- encephalon by stalks that become the optic nerves. Further differentiation of the optic cup gives rise to each of the components of the eye. The lens and cornea arise from the surface ectoderm; the retina, pigment epithelia, and optic nerve from neural ecto- derm; and the vasculature and sclera from paraxial mesoderm. By three months gestation, each of the major anatomical structures is in place. At birth, the axial length of the human eye is about 17 mm (about 74% of adult), and increasing by about 0.16 mm/week. 1 The eye grows nonuniformly; most of its increase in volume is from posterior segment growth. The neonatal corneal surface area is 3/4, and the scleral surface area 1/3 of the emmetropic adult’s. 2 The lens continues to grow after birth, increasing in diameter more than in thickness, resulting in a less spherical and more disk-like adult shape. By 13 years of age, the eye has reached an average axial length of 23 mm, its developmental endpoint. 3 The retina The fovea develops before the peripheral retina; 4–6 yet it is immature at birth. 7,8 It first appears as a bump formed by ganglion cells. Over about the next 25 weeks, foveal ganglion cells and inner nuclear layer cells migrate peripherally, creating the familiar foveal depression at about 15 months. 8,9 Among the many specializations that endow the primate fovea with supreme vision is its peak density of photoreceptors. At birth, the density of foveal photoreceptor cells is a tiny fraction of the adult’s. Peripheral photoreceptor cells migrate toward the fovea from before birth to at least 45 months (longer than the centripetal migration of ganglion cells). As the cones pack together there is a reduction in their diameter; their short, squat inner segments elongate, and the rudimentary stumps of outer segments lengthen into the long, thin appendages of the adult (Fig. 2.1). Since ganglion cells and photoreceptors migrate in opposite directions, extended connecting processes form between the cone pedicles and their cell bodies. Reaching radially as far as 0.4 mm, these specialized axons are the Henle fiber layer, which surrounds the fovea by 2.5 mm in the adult. 10 The human fovea remains immature, even at six to eight months postpartum. Cell morphology and cell density take 15 months to approach maturity, and it may be four years before the retina is largely adult-like. 11 This time course is consist- ent with some aspects of vision development measured experimentally. The retina contains seven principal cell types, each with its own circuitry, organized into layers. The different cell types derive from progenitor cells in the inner layer of the optic cup. Progenitor cells generate different retinal cell types, right up to its final division, raising the following question “What influences the fate of retinal progenitor cells?” The type of cell a progenitor may become follows a temporal order during development that is preserved between species. 12–14 Ganglion cells develop first, followed in overlapping phases by horizontal cells, cones, amacrine cells, rods, bipolar cells, and Müller cells (Fig. 2.2). 15 The acquisition and loss of their ability to differentiate into particular cell types suggests that extrinsic factors serially bias progenitor cells to a particular fate. However, in vitro experiments have not confirmed this. Environmental factors can change the proportions of different cells types generated at a particular stage, but they cannot induce the production of cell types inappropriate for that stage. 16,17 Thus, progenitor cells pass through a number of states during which they are only competent at producing a subset of cell types, and the proportions of city types that they produce at each stage is Normal and Abnormal Visual Development CHAPTER 2 Daniel L Adams ch.02.qxd 29/3/04 05:49 AM Page 9
Transcript
9
THE DEVELOPING VISUAL PATHWAYS
Human vision requires precise collaboration of diverse structures.From the eye to the cerebral cortex, the components mature inparallel, each influencing the development of the whole.
Some developmental processes follow an innate plan that isprogrammed using molecular cues forming “hard-wired” neuralcircuits. Others are controlled by the neuronal activity within thesystem itself that arises spontaneously, or from visual stimulation.Thus, the anatomical configuration of the visual system is sculptedby both nature and nurture.
In a developing individual, visual experience adjusts the neuralstructures such that they best represent the world they areexposed to. Combined with the innate, “hard-wired,” plan, thisproduces an efficient visual system because only elements thatfunction appropriately are maintained: “use it or loose it.”
Reliance on visual experience makes the system vulnerable: afault during development may be detrimental. With anomalousvisual experience, the system develops abnormally. Thus, theprocesses that normally generate an efficient visual system cancause abnormal development.
An example is monocular deprivation. Here, a problem thatobscures vision in one eye, like a congenital cataract, causesabnormal visual development. The visual system allots less corticaltissue to the deprived eye, generating a permanent visual loss thatpersists after removal of the cataract.
The interdependent elements of the visual system must alldevelop appropriately. A fault at any point from eye to brain canhave effects on the whole system. The normal development ofeach component will be discussed individually.
The eyeThe eyes differentiate early from the neural plate. They first appearas the optic pits by the fifth week of gestation, and then extendfrom the neural tube to form the optic vesicles. These sphericalpouches invaginate to form the optic cups, attached to the pros-encephalon by stalks that become the optic nerves. Furtherdifferentiation of the optic cup gives rise to each of the componentsof the eye. The lens and cornea arise from the surface ectoderm;the retina, pigment epithelia, and optic nerve from neural ecto-derm; and the vasculature and sclera from paraxial mesoderm. Bythree months gestation, each of the major anatomical structures isin place.
At birth, the axial length of the human eye is about 17 mm(about 74% of adult), and increasing by about 0.16 mm/week.1 Theeye grows nonuniformly; most of its increase in volume is fromposterior segment growth. The neonatal corneal surface area is 3/4,and the scleral surface area 1/3 of the emmetropic adult’s.2 Thelens continues to grow after birth, increasing in diameter more than
in thickness, resulting in a less spherical and more disk-like adultshape. By 13 years of age, the eye has reached an average axiallength of 23 mm, its developmental endpoint.3
The retinaThe fovea develops before the peripheral retina;4–6 yet it isimmature at birth.7,8 It first appears as a bump formed byganglion cells. Over about the next 25 weeks, foveal ganglioncells and inner nuclear layer cells migrate peripherally, creatingthe familiar foveal depression at about 15 months.8,9
Among the many specializations that endow the primate foveawith supreme vision is its peak density of photoreceptors. Atbirth, the density of foveal photoreceptor cells is a tiny fractionof the adult’s. Peripheral photoreceptor cells migrate toward thefovea from before birth to at least 45 months (longer than thecentripetal migration of ganglion cells). As the cones packtogether there is a reduction in their diameter; their short, squatinner segments elongate, and the rudimentary stumps of outersegments lengthen into the long, thin appendages of the adult(Fig. 2.1). Since ganglion cells and photoreceptors migrate inopposite directions, extended connecting processes formbetween the cone pedicles and their cell bodies. Reaching radiallyas far as 0.4 mm, these specialized axons are the Henle fiberlayer, which surrounds the fovea by 2.5 mm in the adult.10
The human fovea remains immature, even at six to eightmonths postpartum. Cell morphology and cell density take 15 months to approach maturity, and it may be four years beforethe retina is largely adult-like.11 This time course is consist-ent with some aspects of vision development measuredexperimentally.
The retina contains seven principal cell types, each with its owncircuitry, organized into layers. The different cell types derive fromprogenitor cells in the inner layer of the optic cup. Progenitor cellsgenerate different retinal cell types, right up to its final division,raising the following question “What influences the fate of retinalprogenitor cells?” The type of cell a progenitor may becomefollows a temporal order during development that is preservedbetween species.12–14 Ganglion cells develop first, followed inoverlapping phases by horizontal cells, cones, amacrine cells, rods,bipolar cells, and Müller cells (Fig. 2.2).15 The acquisition and lossof their ability to differentiate into particular cell types suggeststhat extrinsic factors serially bias progenitor cells to a particularfate. However, in vitro experiments have not confirmed this.Environmental factors can change the proportions of differentcells types generated at a particular stage, but they cannot inducethe production of cell types inappropriate for that stage.16,17 Thus,progenitor cells pass through a number of states during which theyare only competent at producing a subset of cell types, and theproportions of city types that they produce at each stage is
Normal and Abnormal VisualDevelopment
CHAPTER
2Daniel L Adams
ch.02.qxd 29/3/04 05:49 AM Page 9
controlled by environmental factors.18 This is the “competencemodel” of retinal development.19
Once differentiated, retinal neurons must migrate to their adultpositions before they form synapses and to generate the laminarstructure of the retina. This occurs in two stages:
Stage one brings cells imprecisely into bands at roughly theappropriate depths and is moderated by the attractive/repulsiveadhesion properties of the cells. Mutations in genes encodingadhesion molecules or integrins disrupt retinal lamination.20–22
Stage two more precisely organizes cells into uniform mosaics,distributed tangentially, and with perfect laminar distribution. Inthe rat, at birth the horizontal cells have migrated to within a~50-�m-deep sheet, but by day six, they form a regularly spacedmonolayer.23,24 The mechanisms of the second stage may work bymaintaining constant distances between cells, perhaps byminimizing dendritic overlap. If one cell is removed from, or cellsare added to, a developing retinal mosaic, the others shuffle overto regularize the mozaic.25
The chiasmIn primates, ganglion cell axions enter the optic stalk at about sixweeks of gestation. When they reach the optic chiasm they eithercross or remain ipsilateral. Their decision is influenced by (amongother factors26) adhesions molecules and pathway markersl,27
differential gene expression;28,29 and chiasmal template neurons.30,31
These mechanisms guide axons using attractive and repulsivemolecules.32 One such family of molecules is the “slits.” These arethought to govern where the chiasm forms by defining a restrictingcorridor. Disrupting slit expression produces a large, more anterior,secondary chiasm and prevents retinal ganglion cell (RGC) axonsfrom finding their way into the appropriate optic tract.33
The zinc finger transcription factor, Zic2, is expressed inipsilaterally projecting RGCs during their growth from theventrotemporal retina to the chiasm.34 Zic2 regulates RGC axonrepulsion by cues at the chiasmal midline. The proportion of RGCaxons that cross is related to the size of the animal’s binocularvisual field. Retinal Zic2 levels correlate with the animal’s degreeof binocular vision, suggesting Zic2 is an evolutionarily conserveddeterminant of ipsilateral projection.
In primates, the RGCs in nasal and temporal retina are directedto the contralateral and ipsilateral hemispheres respectively, exceptfor a 5° wedge along the vertical meridian (with the overlapincreasing with vertical distance from the fovea) where ganglioncells project to either hemisphere.35 The normal decussationpattern is disrupted in albinos36 and rarely in otherwise normalprimates37 (see Chapter 45).
Myelination of the optic nerve begins only after all RGC axonshave reached the geniculate body (fifth month in humans) andcontinues into early childhood in a brain-to-eye direction, stoppingat the lamina cribrosa.38 Occasionally, it proceeds into the retina,where it appears as white streaks in the nerve fiber layer.39 Sucherrant myelination is normally benign, but rarely it can beassociated with a visual deficit.40,41
Retinogeniculate projectionsAbout 90% of primate RGCs project to the lateral geniculatebody (LGB), the remainder mostly go to the pretectum and thesuperior colliculus.42 The LGB contains six principal layersspecified by their eye of input and by their cell type (Fig. 2.3).Four of the principal layers (two for each eye) are made up ofsmall, parvocellular (P) cells, and two (one for each eye) containlarge, magnocellular (M) cells. Tiny koniocellular (K) cellsconstitute a third class that occupy the leaflets between the sixprincipal layers.43
The layers of the LGB are present at birth. Their developmentexemplifies the interaction between activity-dependent and hard-
EPIDEMIOLOGY, GROWTH AND DEVELOPMENT
SECTION
1
10
Cone pedicle
Fiber of Henle
Newborn
Innersegment
Outersegment
45 monthpostpartum
5 µm
22 weeks gestation
Fig. 2.1 A schematic drawing showing the stages of development of ahuman foveal cone (left to right) at 22, 24 to 26, and 34 to 36 weeksgestation, newborn, and 15 and 45 months postpartum. The innersegment is present before birth, while the outer segment develops mainlypostnatally, being little more than a stump at birth. The cone pedicle andthe fiber of Henle are present before birth. All four structures undergoextreme postnatal thinning and elongation. Adapted from Hendricksonand Yuodelis.11
Ganglion
Horizontal
Cones
Amacrine
Rods
Bipolar
Müller glial
11 13 16 Birth
Retinogenesis
3 6 9 11days
Embryonic Postnatal
Fig. 2.2 Retinal neurogenesis proceeds in a characteristic sequence.Ganglion cells and horizontal cells differentiate first, followed inoverlapping phases by cones, amacrine cells, rods, bipolar cells, andMüller glial cells. Curves represent the relative proportions of cellsdifferentiating at each stage, not their absolute numbers. The time scalerefers to mouse development. Adapted from Marquardt and Gruss.19
ch.02.qxd 29/3/04 05:49 AM Page 10
wired mechanisms in the formation of the visual pathways. In thehuman, optic tract fibers begin to reach the LGB by about the11th week. Initially, left and right eye afferents are intermingledover the prospective left and right eye layers.44 Between weeks 14and 30, a time corresponding to the formation of eye-specificlayers in the geniculate body, the population of RGC axons reducesfrom 3.5 million to about 1 million.45,46 Perhaps the purpose of thiscell loss is to generate eye-specific layers by eliminating inap-propriately connected axons: “selective elimination.”45
Selective trimming of axon terminal arbors also segregatesbinocular inputs to the LGB. In the cat, axons innervating theLGB grow promiscuous side-branches that contact cells in both leftand right eye columns. By birth, the side-branches contacting theinappropriate layer for their eye are withdrawn, leaving a precisesegregation of inputs.46,47 However, selective elimination ofwhole retinal afferents (rather than single branches) can accountfor the segregation of binocular afferents in the primate LGB.48
In contrast to the laminar segregation by eye, development ofthe M and P layers of the LGB does not employ selectiveelimination of afferent arbors. RGCs become M and P types soonafter their final mitosis.49 The P-type retinal afferents reach the
geniculate first and innervate the medial segment that will laterdevelop into the four P layers. The M-type retinal afferents arrivelater and innervate the lateral segment that will become the twolayers of the M division. Thus, retinal afferents innervate theirappropriate presumptive M or P divisions exclusively, suggestingselective targeting rather than corrective selective elimination.50
Thus, the segregation of the geniculate into magnocellular andparvocellular layers is less dependent on visual experience thanits segregation by eye.
Geniculocortical connectionsMost geniculate cells project to layer 4 of the striate cortex,where inputs from the two eyes differentially activate singlecells–”ocular dominance.”51 As an electrode is advanced parallel tothe cortical layers, the ocular dominance of cells alternatesbetween the left and right eyes. Small lesions of single (monocular)layers of the LGB cause degeneration of terminals in layer 4 ofthe striate cortex in a stripy pattern.52 These 300- to 400-�m-wide stripes of left- and right-eye inputs form a mosaic ofdiscreet columns. The complete pattern of ocular dominancecolumns can be visualized by radioactive tracer injected into oneeye.53 The tracer is taken up by RGCs and transported to layer 4of the striate cortex (Fig 2.4).
Formation of the ocular dominance column pattern cannot bedependent on visual experience because (at least in the macaque)it is adult-like at birth.54,55 However, this does not necessarilymean that it forms independently of neuronal activity. It has longbeen held that the geniculocortical afferents are initiallyintermingled and are segregated into ocular dominance columnsunder the influence of retinal activity.56–58 Thus, pharmacologicalblockage of retinal activity abolishes column formation in the cat.57
The spontaneous waves of neuronal activity that roll across eachretina in utero could play a role in column segregation bygenerating firing patterns in the RGCs that are spatially correlatedwithin, but not between, each eye.59 This suggests that cellswith synchronous activity are segregated into a single oculardominance column, i.e., cells that “fire together, wire together.”60
However, ferrets binocularly enucleated before their genicu-locortical afferents arrived at the striate cortex form normal
Normal and Abnormal Visual Development
CHAPTER
2
11
Fig. 2.3 Macaque LGN following injection of a radioactive tracer ([3H]proline) into the right (contralateral) eye. Layers containing the tracerappear bright. The six principal layers are monocular in macaque (andhuman) and follows the sequence: ipsi-contra-contra-ipsi-contra-ipsi.Layers 1 and 2 are magnocellular, and layers 3–6 are parvocellular.Koniocellular layers are situated between the six principal layers (not seein this section). M, magnocellular; P, parvocellular; c, contralateral; i, ipsilateral.
Fig. 2.4 Macaque monkey left striate cortex following injection of aradioactive tracer ([3H] proline) into the left eye. The tracer appears bright.The tissue has been dissected from the rest of the brain, unfolded, andflattened to show the entire striate cortex. The mosaic of oculardominance columns is visible because the tracer was transported to onlythose columns belonging to the injected eye. The oval in the center is therepresentation of the blind spot—a monocular region of the visual field.
ch.02.qxd 29/3/04 05:49 AM Page 11
ocular dominance columns.61 Thus, retinal activity cannot be aprerequisite for columnar segregation in this species. The forma-tion of the ocular dominance column pattern may rely on intrinsicsignals, e.g., molecular cues on thalamic axons, on cortical cells, oron both.62
Extrastriate cortical areasCells in the striate cortex project to multiple extrastriate visualareas, which form an interconnected hierarchy reaching into theparietal and temporal lobes. Many regions in this network aredefined as single visual areas by their retinotopic organization, cellselectivity, and/or unique pattern of connections with otherregions.63 It is thought that different areas process differentvisual modalities, e.g., motion, color, and form.64
The development of extrastriate cortical areas has beeninvestigated by the early removal of cortical tissue, before anyinputs have arrived at the neocortex:65 If cortical areas form inpredetermined regions, subsequent mapping of visual areas inthese animals in adulthood would show a reduced number of areas.However, they show a full compliment of cortical areas, squeezedonto a reduced area of cortex. Thus the primitive neocortex is anunspecialized substrate whose subdivision into areas occurs inunison.
NORMAL VISUAL DEVELOPMENT
To study visual development, it is important to know about thevision of the newborn. Newborn humans can see, they prefer tolook at faces,66,67 they can discriminate between mouth openingand tongue protrusion, and rapidly imitate either.68 They initiallyfixate simple high-contrast patterns (like their mother’s hairline),and are later attracted to more subtle features (like theirmother’s eyes).
Measurement techniquesSince infants are unable to report what they see, adult visionmeasurement techniques must be adapted and other indicators of“seeing” used.69,70 The three most commonly used techniques willbe described:
Preferential lookingGiven a choice of looking at a striped grating or a uniform field, aninfant will prefer to fix the grating.71–73 A hidden observer guesseswhich of the stimuli contains a grating based on the infant’s fixationpattern.74 As less visible gratings are presented, the observer makesmore incorrect inferences. The visual threshold is defined by thegrating stimulus that generates only 75% correct inferences fromthe observer. While this technique has been successfully used toasses infant visual development,74–76 its reliability is dependenton many trials and it is not always possible to hold the infant’sattention for the necessary time.77
Visually evoked potentialsThe visually evoked potentials (VEP) technique measures electricalactivity directly from the scalp using surface electrodes.78,79 Visualstimulation produces a stereotypical wave whose amplitude andtiming can be measured. Repeated responses are recorded andaveraged to improve signal-to-noise ratio. A “transient” VEP isrecorded in response to a single event, e.g., a flashed stimulus,80
and a “steady-state” VEP is a continuous standing wave patternproduced by a rapidly repeating stimulus.81
The raw VEP signal must be analyzed to estimate acuity. Asimple method is to compile a set of transient VEPs using a singlehigh-contrast stimulus, and measure the Snellen acuity in anemmetropic adult with various degrees of optical blur. The infant’sacuity can then be estimated by comparing their VEP to theblurred adult set. The acuity of the infant is presumed to be equalto the acuity of the adult whose vision was blurred such that theirVEP signals were most similar.80,82 This method relies on theunlikely assumption that infant and adult VEPs are equivalent.An alternative method is to present finer and finer grating stimuliuntil a transient VEP is no longer measurable above noise,83 or toextrapolate to zero response amplitude, using either thetransient84 or the steady-state VEP.85 Extrapolation to zero isused because the VEP signal is inherently noisy, making it moredifficult to define the exact point where a small signal disappears.
Optokinetic nystagmusA wide field-drifting stimulus generates optokinetic nystagmusthat can be used to estimate infant visual acuity because it is onlygenerated by resolvable stimuli and it is easily observed. Theearliest investigation used a stimulus attached to a metronomewand;86 later experiments employed scrolls of paper, printedwith gratings and streamed over the infants visual field by a handcrank.73,87 Eye movements were observed or measured withelectro-oculograms.88
Visual acuityVisual acuity (“grating acuity,” “resolution acuity”) is a measure ofthe finest feature detectable by an observer. It can be described asthe visual angle subtended by a single stripe element (minutes perstripe), or more formally as its reciprocal (cycles per degree), i.e.,the threshold spatial frequency of a 100% contrast square-wavegrating.89 In normal adults, resolution acuity is equal to about 1 min/stripe, or 30 cycles/degree. By setting this value to beequivalent to the standard Snellen acuity of 20/20, it is possibleto roughly convert between the two scales.
In classic studies using optokinetic nystagmus (OKN), 93 of 100infants aged 80 minutes to 5 days responded to a 0.56 cycles/degreegrating moving at 8.5°/sec, but none responded to a 0.19 cycles/degree grating moving at the same speed.87 Using a greater rangeof grating sizes, a “large percentage” of another 100 newbornsresponded to a 0.25 cycles/degree grating.90 Nearly all the infantstested had at least 20/400, and some 20/300, Snellen equivalentvision. Others found lower values for newborn acuity,91 but thisrange is consistent with most investigations of zero- to three-week-old infant acuity, using OKN73,88 and PL.75
Natural variation, nonstandardized testing techniques, differentviewing distances, and illuminations produced differences inmeasured acuity between studies. To overcome this, thestandardized Teller Acuity Card system was devised.92,93 Usingthese cards, 140 infants showed a mean acuity at one week of 0.9(± ~0.5 SD) cycles/degree.94 This data, along with other measuresof the development of acuity using PL, is shown in Fig. 2.5.
A comparison of PL with VEP data shows that the VEPtechnique gives acuity values about 1 to 2 octaves better than PLmeasures at all ages (an octave is a doubling of acuity). This couldbe due to the stationary stimuli used in PL studies, whereas VEPstudies use temporally modulated gratings that may produce alower acuity threshold. If checkerboard stimuli are used in VEPstudies, the amplitude of the signal shows a peak at a particularcheck size. If the location of this peak is used instead of the VEPamplitude, the two acuity measures agree.95 When VEPs and FPL
EPIDEMIOLOGY, GROWTH AND DEVELOPMENT
SECTION
1
12
ch.02.qxd 29/3/04 05:49 AM Page 12
were measured in the same infants, VEP signals could be detectedfor spatial patterns that were below threshold for behavioralmeasures. This could be due to the signal averaging used in theVEP technique.96 It seems that the increased signal-to-noise ratiogenerated by averaging in VEP studies is of benefit to theexperimenter but not the visual system! If VEP latency was usedinstead of amplitude, comparable scores could be generated withthe two techniques.
Contrast sensitivityA more complete evaluation of the spatial performance of thevisual system can be gained by measuring many contrast thresholdsover a range of spatial frequency. This produces a graph ofthreshold contrast versus spatial frequency–”contrast sensitivityfunction.”97,98 The conventionally measured grating acuity is thenrepresented by the abscissa of the x axis (100% contrast). Theadult curve has a peak value at 3–5 cycles/degree and a decline insensitivity at both lower and higher spatial frequencies.
Overall contrast sensitivity of the infant is about 10 timeshigher than the adult but infants are relatively more sensitive inthe low-frequency region. This reduced “low-frequency cut” wasdemonstrated in a PL study of infants from age 5 to 12 weeks.99
The low-frequency cut was smallest in the 5-week group, morepronounced by 12 weeks.
The absence of a low-frequency cut in the 5-week group, andits relatively smaller size in the older groups, suggests that theundeveloped visual pathway might be relatively well suited totransmit coarse features that do not require the high-resolutioncomponents of the retina, or it could be central in origin. Inhibitorycortical connections may tune cortical cells to higher spatialfrequencies;100,101 and perhaps these are underdeveloped in infants.It could be artifactual, due to the reduced number of cycles visiblein the low-frequency stimuli,102–104 or perhaps infants just prefer tofixate lower spatial frequencies. The effect has been verified by
further PL investigations (Fig. 2.6) but the origins of changes in thecontrast sensitivity function over the first 3 months of infancywould have to be determined with other techniques.
The maturation of contrast sensitivity has been studied in theinfant using the steady-state “sweep VEP,”105–107 where the spatialfrequency of the stimulus is swept over a range of values duringrecording, while holding contrast constant. These findingssubstantiated the main PL findings and narrowed the search foranatomical correlates. The absence of low-frequency attenuation inyoung infants was reproduced, and its time course clarified. Up to9 weeks of age, the contrast sensitivity function shows an increasein sensitivity at all spatial frequencies. Thence, sensitivity increasesare restricted to the higher frequency domain, indicating animprovement in spatial resolution, rather than sensitivity per se.Thus, the development of low-spatial-frequency vision follows atime-course shorter than that of high spatial frequencies. Giventhat high spatial frequencies are detected by the fovea, whichdevelops more slowly than the periphery,7 it is likely that the slowincrease in high-spatial-frequency sensitivity is a result of theprolonged development of the fovea. Likewise, since low-spatial-frequency sensitivity is unaffected by exclusion of the fovea,108 itfollows that the early relative sensitivity to low spatialfrequencies is due to the relatively advanced maturation ofperipheral photoreceptors.
An infant monkey, operant trained to indicate with a lever pushthe screen that contains a grating, shows that the macaque visualsystem is very similar to the human’s, but it develops muchfaster.109 Infant humans and monkeys show the same depressionof sensitivity compared to adults. The higher relative sensitivity
Normal and Abnormal Visual Development
CHAPTER
2
13
4
8
1
2
0.5
0 1 4 6 8 10 12
Gwiada et al
Birch & Hale
Van Hof-van Duin and Mohn
Allen
Mayer and Dobson
McDonald et al
Courage and Adams
Acu
ity (c
ycle
s/d
eg)
Age (months)
Fig. 2.5 Representative examples of data showing the development ofbinocular grating acuity in normal infants tested with the PLprocedure.70,94,241–245 From Dobson.240
30
100
10
3
300
1000
.030.1 1.0 3.0 10 30
Adult
3 months
2 months
1 month
Con
tras
t se
nsiti
vity
Spatial frequency (c/deg)
Fig. 2.6 Average contrast sensitivity functions for 1-, 2-, and 3-month-olds and an adult obtained using an FLP procedure. The solid linerepresents data obtained with the infant apparatus; the dashed portionrepresents typical high-frequency data for an adult under similarconditions. Data from Banks and Salapatek.76
ch.02.qxd 29/3/04 05:49 AM Page 13
to low spatial frequencies of human infants is also apparent, as isthe differential developmental time-course for low and highspatial frequencies.110 Young monkeys’ foveal contrast sensitivityis similar to the near periphery but develops to a greater degreeduring maturation.111 Thus, the developmental time-courses ofhigh- and low-spatial-frequency sensitivity matched those of thefovea and peripheral retina, respectively.
To compare the development of grating acuity and contrastsensitivity in the central and peripheral visual field of the humaninfant, sweep VEPs were measured in infants ranging from 10 to39 weeks of age.112 By testing the central and peripheral fieldssimultaneously, each at a different temporal frequency, it wasfound that peripheral acuity reached adult values by 26 weekswhile central acuity did not reach asymptosy up to 33 weeks. Thehuman infant’s fixation sensitivity was better than its peripheralvision at all ages. However, this could be due to the infant usingan eccentric fixation strategy, rather like the one an adult mightuse to view a distant star at night.
Photoreceptor development is probably not the only limiter ofmaturation of contrast sensitivity.113 Optical influences are modestbecause the neonatal media are clear, and accommodation does notaffect acuity over distances between 30 and 150 cm.114 To gaugeany influence the LGB and striate cortex may have on contrastsensitivity is not straightforward because they receive signalsalready filtered by the immature retina.
Binocular visionPrimates and carnivores perceive depth from their binocular viewof the world: stereopsis. For stereopsis to develop, eye movementsmust bring both foveas at the same point in 3-D space and theeyes must move conjugately to maintain binocular fusion duringvision. When the images from the eyes are continuously fused,the disparities are analyzed by cortical cells for the perception ofdepth. Stereopsis is robust: stereograms can be perceived whenone eye’s image is blurred;115 likewise, it can develop in infantsdespite significant anisometropia.116
Stereopsis is the result of a neuronal calculation, so it requiresspecialized stimuli to isolate it from other visual cues. Clinically,stereopsis is often tested with the Titmus test. The patient wearspolarized spectacles with the plane of polarization at right anglesfor each eye. Test images consist of superimposed stereo-pairs,presented differently to each eye by a polarized filter layer. Oneexample is of a large housefly, which stands out from the page toan observer with stereopsis. Others are circles and animal picturesat different depth planes. While these tests are adequate forassessing the presence or absence of stereopsis in children, randomdot stereograms117 contain a form that can only be seen withstereopsis. Viewed monocularly or by the stereoblind, a randomdot stereogram appears as a flat field of noise. When viewedstereoscopically, pictures appear in front of, or behind the planeof the page.
By about two months of age, some infants, tested with randomdot stereograms, apparently discriminated disparity when testedwith PL and a habituation recovery test.118 Computers madedynamic random dot stereograms more amenable. Using FPL,infants were presented with a square stimulus, defined by stereoalone, that drifted to the right or left. The hidden observerguessed the direction of the stimulus by the infant’s behavior andeye movements.119 No significant difference from chance wasmeasured in the observer’s direction guesses for infants up to 3.5 months of age, and even later (up to 6 months) in a longi-tudinal experiment.120
Random dot stereograms are suited for use with VEPs becausethey are perceived only if stereopsis is present. Any modulation ofthe VEP signal at the same frequency as the stimulus is indicativeof stereopsis. Random dot stimuli that oscillate in depth(stereograms) or that counterphase in one eye (correlograms)evoke large potentials, enabling reliable determination ofstereopsis in infants.121–123 Stereo stimuli evoked responses ininfants at 8 to 20 weeks.124,125
Stereoacuity improves until about 24 months, when itapproaches adult levels126,127; some improvement is from increasedinterocular distance but most is due to development of the centralvisual pathways. Disparity-tuned binocular neurons have beenfound in the striate cortex in both the cat and monkey.128,129
Although these cells signal the disparity of stimuli, their responsesdo not necessarily correlate with depth perception.130,131 Cells inthe monkey striate cortex can be tuned to disparity by the sixthpostnatal day,132 several weeks before the onset of stereopsis.Furthermore, infant monkeys had an adult proportion ofdisparity-tuned cells and their ocular dominance histograms areadult-like.133 The development of stereopsis is not due simply toa proportional increase in the numbers of disparity-tuned cells inthe striate cortex, but to a refinement of their spatial responseproperties and overall responsiveness. The onset of stereopsismay correlate with the refinement of extrastriate visualconnections and increasing populations of disparity-tuned cells inhigher visual areas.
Orientation selectivityCells activated selectively by bars or gratings presented over asmall range of orientations are implicated in form vision.134
Recordings from striate cortex cells of newborn kittens showedthat no neurons were adult-like in their responses to orientedcontours.135 Orientation-selective cells have been found invisually inexperienced macaque striate cortex at three weeks.136
Even at this early stage, the cells are organized into adult-likecolumns that are in register with ocular dominance columns.137
The relationship between orientation selectivity and visualexperience has been investigated by raising animals in an artificialenvironment that exposed them to a restricted range of contourorientations: “stripe rearing.”138,139 Stripe-reared cats were firstshown to have a larger than normal proportion of cells tuned tothe orientation they were exposed to most. Despite the dramaticchanges in the distributions of orientation preference of corticalcells, only modest specific behavioral deficits could be measured,incommensurate with the magnitude of the physiological effects.140
This inconsistency led others to re-examine the phenomenon. Atfirst, no effect was found;141 later, a small effect was described.142
The inconsistencies in stripe rearing experiments were probablydue to different techniques and sampling methods. The overalleffect, though probably real, was certainly not as striking as it wasfirst described.
Motion perceptionMotion information is crucial to many visual and motor functions,for example, the encoding of depth through parallax, estimatingtrajectories, segmenting figures from backgrounds, and controllingposture and eye movements.143 OKN studies show that motionvision develops early in humans.144 However, OKN is a crude andreflexive measure of motion sensitivity that does not require finedirection discrimination.
EPIDEMIOLOGY, GROWTH AND DEVELOPMENT
SECTION
1
14
ch.02.qxd 29/3/04 05:49 AM Page 14
The ability to discriminate opposite directions of motiondevelops at about 10 to 13 weeks.145,146 VEP studies show it withinthe first two months of life.147,148 Finer discriminations of motiondirection have been tested using FPL where infants were presentedwith windows of dots moving in a different direction to thebackground dots.148 The angle between target and backgrounddirections was reduced until preferential looking by the infantwas no longer detectable. By the age of 12 weeks, infants madequite fine discriminations, on the order of 20°, and by 18 weeksthey were down to ~15° These values are still far from those ofadults, who have no trouble making discriminations of less than 1°.
Neural motion perception is thought to result from the activityof direction-selective cells. These are found in layer 4B of thestriate cortex,150 and in many extrastriate cortical areas, mostnotably V5 (MT).151 Electrical microstimulation of V5 cells inthe monkey has shown that their activity correlates directly withthe perception of motion direction and can affects the animal’sdirection discrimination.152 Little is known about the developmentof direction selectivity in area V5. However, single-cell recordingsin the striate cortex of 1-week-old monkeys have shown thatdirection selectivity is absent or very broad. The tuning width was45°–90° by 2 weeks of age, narrowing to an adult-like 30° by 4 to8 weeks.153 Thus, the time-course of direction selectivity in monkeystriate cortex approximately matches that of the psychophysicalmeasurement of direction discrimination in the human infant,counting the four-times slower maturation of the human visualsystem.109
Color visionEarly studies suggest that very young infants are able todiscriminate different colors.154 However, most natural-coloredstimuli also differ in their real and perceived brightness. To testcolor vision exclusively it is necessary to use colors of the sameperceived brightness that can only be discriminated by theirwavelength composition (isoluminant colors). Different individualscan have different isoluminance points, so to eliminate individualvariability, isoluminance points must be measured in everyexperimental subject. In early studies, adult isoluminance pointswere used,155 introducing a luminance confound.156 FPL tech-niques for measuring isoluminance points in infants were soondeveloped.157,158 FPL-derived isoluminance points are unlikely tobe perfectly accurate, so residual luminance artifacts werecamouflaged either by testing over small ranges of luminancedifferences from trial to trial or by dividing the stimuli into anumber of tiles and randomly “jittering” their luminance.
With the luminance confound removed, the infants’ chromaticdiscrimination can be tested, usually with FPL, using patterned(preferred) stimuli, commonly two isoluminant color checks orgratings, versus uniform (nonpreferred) intermediate color stimuli.Thus it was shown that 8-week-old infants can distinguish a redand isoluminant gray square wave grating from a uniformluminance matched stimulus.158 Female infants were used toreduce the probability of a color-blind infant being tested. A few4-week-olds, some 8-week-olds, and all 12-week-olds demon-strated color vision.159–161 Sensitivities to different wavelengthsmay develop at different times, making infants functionallydeuteranopic for a developmental period.160 However, a widerange of normal variation exists in development of spectralsensitivities.161,162
Luminance and chromatic sensitivity are both dependent on thesame photoreceptors: red and green cones.163 If the time-courseof changes in contrast sensitivity are the same for luminance and
chromatic stimuli, it suggests that their neural correlate lies withphotoreceptors development,164 but if their time-course is differ-ent, separate (post-receptor) mechanisms may be responsible foreach.165 To differentiate, it is necessary to measure chromaticcontrast sensitivity in infants over a range of ages. Using VEPs, itwas found that chromatic and luminance contrast sensitivityfunctions at all ages were well described by curves of a commonshape, with developmental changes confined to upward shifts insensitivity and rightward shifts in spatial scale.166–169 Thus, theirpoorer color discriminative ability (like their poorer luminancecontrast sensitivity) can be explained by the smaller percentageof photons caught by their immature photoreceptors.164
ABNORMAL VISUAL DEVELOPMENT:AMBLYOPIA
Definition“Amblyopia” is from the Greek amblyos, blunt, and opia, vision.Albrecht von Graefe defined amblyopia as the condition in whichthe observer saw nothing and the patient very little. This definitionremains valid because it emphasizes an important feature ofamblyopia–that looking into the eye reveals nothing about thedisease itself. Eye examination does reveal factors that causeamblyopia, like cataract, strabismus, and anisometropia. A moreformal definition of amblyopia is visual impairment withoutapparent organic pathology.
Critical periodsThe term “critical period” was first used by Konrad Lorenz in hisstudies of imprinting in birds. It was adopted by Hubel andWiesel to refer to the time when deprivation changes the oculardominance of cells in the striate cortex.170 It falls between 4 and6 weeks in the cat,171 when closure of one eye for 3 days or moreleads to a visual cortex dominated by cells responsive only to theopen eye. Some susceptibility to deprivation persists until about9 months.172 Since Lorenz coined the term, it has taken on amore general use. A critical period can be defined for any func-tion as the time when, if deprived of normal stimulation or unused,its development may be permanently disrupted. Visual criticalperiods begin after the initiation of visual stimulation (eye-openingin cats, birth in primates) and last between weeks and years,depending on the species and visual faculty in question.
Critical periods have been defined for strabismus,173 for thedevelopment of direction selectivity in cat striate cortex,173 andfor orientation selectivity.175 Neurones with higher functions, likebinocularity, have critical periods that end later than those withearlier-processed properties, like ocular dominance. This isexemplified within the striate cortex, where the critical period isover in layer 4 (the input layer) before the other layers.176
Early monocular deprivation is catastrophic to the visualsystem because it affects many critical periods. Critical periodscan be inferred by monocular deprivation after various delays. Inthe monkey, monocular deprivation before 3 months affectsabsolute light sensitivity, between 3 and 6 months sensitivity towavelength and brightness, up to 18 months high-spatial-frequencyvision, and up to 24 months binocular vision.177 Human criticalperiods are less well defined, and can be deduced by studyingchildren with amblyopia following unilateral cataract surgery.178
By comparing children in whom the age of cataract onset andcorrection of vision were known, the human critical period forvisual acuity loss seems much longer than that in cats and monkeys.
Normal and Abnormal Visual Development
CHAPTER
2
15
ch.02.qxd 29/3/04 05:49 AM Page 15
Weeks of deprivation between 6 and 18 months and months ofdeprivation up to age 8 produce permanent visual deficits. Earlycorrection is therefore imperative for visual recovery followingdeprivation.
Amblyopia may recover in adult humans following loss of thenonamblyopic eye. This occurs well beyond the critical period,when no amount of monocular occlusion would result in theinduction of amblyopia. Improvements have been measuredfollowing visual loss of the good eye in teenagers and adults179 anda 65-year-old.180 One year after the loss of their good eye, 20% of254 amblyopes aged 11 years or older had some improvement intheir amblyopic eye,181 and half improved by two or more Snellenlines. At least in a minority of patients, neural plasticity can occurpast the critical period.
Factors other than neural plasticity could also lead to an apparentimprovement in visual acuity. Fixation may become more stableand accommodation accuracy may improve in an amblyopic eyeonce the individual is forced to use it alone. Adult recovery fromamblyopia may help us understand the factors that normally actto restrain plasticity beyond the critical period but childrenshould remain the focus of detection and treatment.181
CausesAmblyopia is caused by abnormal visual experience: mostly bystrabismus, anisometropia, monocular form deprivation, or a com-bination of these. Defining the cause in any particular patient is notalways straightforward because anisometropia and strabismus canarise as a consequence of amblyopia,182 making it difficult to dis-tinguish cause and effect unless a patient is examined early enough.The three most recognized causes are strabismus, anisometropia,and monocular form deprivation.
StrabismusStrabismus is a misalignment of the optic axes resulting from motoror sensory deficits.183 The optic axes may be crossed (esotropic),diverged (exotropic), or vertically misaligned (hyper/hypotropic).Humans are often born with a slight exodeviation, thought torepresent the anatomic positions of the divergent orbits.182 Duringthe first six months, binocular fusion emerges, and a normal infantattains orthotropic vision.183 One to 2% of infants do not developbinocular fusion and acquire strabismus.181 Esotropia is the mostcommon form of childhood strabismus and is often associated withhyperopia. Infants are born hyperopic but generally attainemmetropia, though some remain hyperopic until 2–3 years ormore.186 Some of these infants accommodate to correct theirblurred vision. Normally, accommodation is reflexively accom-panied by a tendency to converge but in hyperopia this reflex cancause the eyes to cross, preventing binocular fusion and resulting ininfantile esotropia.
Strabismics rarely complain of double vision because theysuppress perception from the deviated eye. Some are able toalternate fixation and suppression between their eyes, and theyrarely develop amblyopia. Strabismics who fix constantly with oneeye and suppress its deviated fellow are most at risk of developingamblyopia. In adult monkeys made exotropic surgically, themetabolic activity of one or other set of ocular dominance columnswas found to be locally depressed.187 The retinotopic locations ofdepressed columns corresponded to the locations of suppressionscotomas in human exotropes. This demonstrates that strabismicsuppression reduces the activity of cells in the striate cortex. Theneural mechanism of suppression may be similar to that of normalbinocular rivalry.188
AnisometropiaAnisometropia is an interocular difference in refractive power,often the result of a difference in size or shape of the globes. Aninequality of greater than 2 diopters is potentially amblyogenic ifit persists until the age of 3 or longer.183,189 In humans, about one-third of cases of anisometropia are accompanied by strabismus.Blurred vision in one eye may lead to inaccurate binocular fusion,resulting in a small angle strabismus, but confounding effects ofthe strabismus are difficult to isolate.
Anisometropic amblyopia has been produced in primates bydaily uniocular administration of the cycloplegic, atropine, frombirth to eight months.188 This provides a realistic model of theeffects of anisometropic amblyopia without the confounding factorsthat exist in humans.191–193
Astigmatism occurs when one or more of the refracting surfacesof the eye contain a cylindrical component, resulting in refractivepower that varies at different meridians. Astigmatism is common(30 to 70%) in the first two years of life.194–197 Early astigmatismmay not to have any detrimental affect on visual developmentbut, if persistent, it is a risk factor for amblyopia.198
Astigmatism may be responsible for a form of amblyopia that isorientation specific–”meridional” amblyopia,199 where sub-populations of cortical cells are selectively affected according totheir visual response properties. In humans, the angle andmagnitude of meridional amblyopia correlate well with that ofastigmatism.200 Adult monkeys, raised with fixed orientationcylindrical lenses to imitate binocular astigmatism, have shownorientation-specific acuity deficits.190,201
Monocular form deprivationIt is not the lack of light that causes amblyopia, but the lack ofa sharp image. Monocular blur (anisometropia) is a form ofmonocular form deprivation in an eye with clear optics. Formdeprivation can also be caused by light scattering from imper-fections in the optical components of the eye. Cataracts are acommon cause. Surgery for early cataracts is urgent because, atthe peak of the critical period, as little as 2 weeks of deprivationcan initiate amblyopia. However, the operated, aphakic eye is farfrom normal: accommodation is abolished, so focus is fixed at asingle distance and anisometropia occurs at some fixationdistances. Aggressive patching following cataract surgery mayenable some recovery of vision, but it rarely prevents amblyopiaentirely.202–205
Suturing the lids of one eye of neonatal cats causes profoundamblyopia reliably but nonspecifically, because it blocks allmodalities of vision. Monkeys raised with one of three differentstrengths of diffuser spectacle lenses in front of one eye and aclear zero-powered lens in front of the fellow eye were found, inadulthood, to have a close correspondence between the magni-tude of the amblyopia and the reduction in retinal image contrastproduced by the diffuser lenses.206 Thus, the depth of non-strabismic amblyopia is strongly influenced by the degree ofretinal image degradation early in life.
ClassificationPatients are usually classified as strabismic or anisometropicamblyopes based on the symptoms at the time of study. However,strabismus and anisometropia can also arise as a consequence ofamblyopia, making the classification of amblyopia at best difficult.Those amblyopic from an isolated cataract are an exception; it ispresumed that monocular deprivation is the sole cause.
EPIDEMIOLOGY, GROWTH AND DEVELOPMENT
SECTION
1
16
ch.02.qxd 29/3/04 05:49 AM Page 16
A fundamental distinction exists between strabismic and otherforms of amblyopia. Anisometropic and deprivation amblyopiasare caused by an optical degradation of one retinal image, but instrabismic amblyopia both retinal images are initially perfect.
It has been proposed that there are distinct patterns of visualdeficits in strabismic and anisometropic amblyopes,207,208 but thishas been based on studies of few patients.209–212 Recently, 427amblyopes between the ages of 8 and 40 were classified by oculardeviation, surgical history, refractive errors, eccentric fixation, anddeprivation history, and compared with 68 controls.213 Measures ofacuity, contrast sensitivity, and binocular and stereo vision wereundertaken. Three patterns of visual loss corresponding roughlyto traditional classifications based on the associated condition werefound: strabismics, anisometropes, and strabismic anisometropes.Deprivational amblyopes had functional deficits distinguishablefrom anisometropes (Fig. 2.7). Thus, two developmental anomaliescould account for the patterns of visual loss in amblyopia–poorimage formation in one eye (anisometropes) and a loss of binocularfunction (strabismics). A combination of these factors producesthe third, worst affected group (strabismic amblyopes).
Visual deficitsClinically, amblyopia is characterized by a significant reduction ofSnellen visual acuity that does not correct with refraction. How-ever, Snellen acuity is a general measure of visual function. If morespecific characteristics of vision are examined, a more precisepicture of amblyopia can be gained. Comparing grating acuity,Vernier acuity, and Snellen acuity in strabismic and anisometropicamblyopes shows that grating acuity is more reduced inanisometropic than strabismic amblyopia.214–218 Conversely,contrast sensitivity is more elevated in strabismic amblyopes thanthose with preserved binocular vision (including controls) butreduced below normal in anisometropic amblyopes.213
Interference effects, characteristics of normal spatial vision,manifest as a reduced discrimination of closely spaced stimuli, e.g.,orientation,219 stereoacuity,220 and Vernier acuity.219 Spatialinterference, or “crowding,” is elevated in amblyopic eyes,221,222 soamblyopes’ poor performance at Snellen charts organized in rows
may be improved by using single optotypes.221,223 Visual measuresadversely affected by crowding all rely on hyperacuity; i.e., they arelimited by cortical processing rather than the spatial resolution offoveal cones.224,225 Since amblyopia is a cortical deficit, diminishedhyperacuity and the increased effects of visual crowding are typicalin amblyopic individuals.
The acuity and sensitivity deficits in amblyopia could be theresult of changes in striate cortex neurons,226–228 but other deficitsare less easily accounted for by “low-level” visual neuron activity. Ifstrabismic amblyopes are asked to count highly visible featuresthat vary in number, or change orientation in briefly presentedstimuli, they systematically undercount. This suggests a limit tothe amount of information that the amblyopic visual system canattend to individually.229 Cueing the observer to the relevant partof the display improved performance in amblyopes and normalsalike, suggesting that the amblyopic deficit was not the result ofreduced spatial attention. It is unlikely that this “high-level”visual processing is in the striate cortex; but probably reflectsunreliable signals reaching higher visual areas from the striatecortex.
Anatomical correlatesThe balance of left- and right-eye cells in the cat striate cortex canbe tilted in favor of one or the other eye by manipulating earlyvisual experience. Immediately after newborn kittens open theireyes, a few weeks of monocular eye-lid closure result in a paucityof cortical cells responding to the closed eye.230,231 This regimeproduced monocular deprivation similar to a congenital cataract inhumans. In macaques, the changes are accompanied by a change inthe relative widths of the ocular dominance columns.232,233 Thenormal pattern of ocular dominance columns, roughly equal inwidth, is remodeled so that the columns belonging to the suturedeye shrink and the space is taken by the expanded columns of thenondeprived eye (Fig. 2.8).
This anatomical evidence of postnatal remodeling of oculardominance columns in monocular deprivation suggests thatamblyopia could be caused by a lack of striate cortex devoted tothe deprived eye. However, other forms of amblyopia are not
Fig. 2.7 Eleven clinically defined categories ofamblyopia in the study of McKee et al.211 Themean position of each is plotted against “acuity”and “sensitivity,” calculated from a number oftests. The normal, strabismic, and anisometropicobservers fall into different regions of the two-factor space. The strabismic amblyopes appear torepresent a mixture of the strabismic andanisometropic categories. Error bars represent 1SEM along the principal axes of each category’selliptical distribution.
ch.02.qxd 29/3/04 05:49 AM Page 17
associated with differential column shrinkage. The oculardominance columns of a human strabismic and an anisometropicamblyope were to shown to have the same width for botheyes234,235 and a naturally occurring anisometropic amblyopicmacaque also had normal and equal width ocular dominancecolumns.236 Thus, column shrinkage does not necessarily have acausal relationship to amblyopia; nevertheless, column shrinkagewithout amblyopia has never been described.
A fMRI study showed a biased share of cortical territory infavor of the nonamblyopic eye in strabismic, anisometropic, andstrabismic-anisometropic amblyopes whose visual deficitdeveloped during infancy, but no effect if the deficit developedafter two years of age.237 This finding contradicts the histologicalstudies in humans and animals. Perhaps the humans and animalsstudied histologically had late-onset amblyopia, occurring afterremodeling of the columns was possible. The resolution of fMRI islimited to about 0.5 mm, about the width of a single oculardominance column, making it difficult to resolve columns, let alonemeasure the subtle changes that they may undergo in amblyopia.The spatial resolution of histological tissue is not limited.
Anatomical tracer injections have shown that neurons in thestriate cortex of strabismics have abnormal wiring. At birth,intralaminar horizontal connections exist between neighboringocular dominance columns. This pattern of horizontal fibersnormally persists into adulthood,238 but if strabismus is inducedduring the critical period, there is a change in the horizontalnetwork: projections between left- and right-eye columns arereduced, leaving only fibers that connect cells activated by thesame eye.239
Amblyopia is caused by anatomical and functional changes inthe brain. They have been observed in the striate cortex, but it isunlikely that this is the only region altered; it is merely the beststudied. Further investigations of the anatomical wiring andphysiological properties of neurones in amblyopic and normalanimals and humans may tell us more about the mechanisms that cause amblyopia. One hopes that such knowledge willfacilitate new approaches for the treatment and prevention of thedisease.
EPIDEMIOLOGY, GROWTH AND DEVELOPMENT
SECTION
1
18
Fig. 2.8 The ocular dominance column pattern of a macaque, followingearly monocular eye-lid suture to simulate congenital cataract. Thecolumns belonging to the deprived eye appear shrunken and reduced tosmall islands, while those of the unaffected eye have expanded theirterritory.
REFERENCES
1. Fledelius HC, Christensen AC. Reappraisal of the human oculargrowth curve in fetal life, infancy, and early childhood. Br JOphthalmol 1996; 80: 918–21.
2. Swan KC, Wilkins JH. Extraocular muscle surgery in earlyinfancy–anatomical factors. J Pediatr Ophthalmol Strabismus1984; 21: 44–9.
3. Sorsby A, Leary GA. A longitudinal study of refraction and itscomponents during growth. Spec Rec Ser Med Res Counc (GB)1969; 309: 1–41.
4. Rapaport DH, Stone J. The site of commencement of maturationin mammalian retina: observations in the cat. Brain Res 1982; 281:273–9.
5. Okada M, Erickson A, Hendrickson A. Light and electron micro-scopic analysis of synaptic development in Macaca monkey retinaas detected by immunocytochemical labeling for the synapticvesicle protein, SV2. J Comp Neurol 1994; 339: 535–58.
6. La Vail MM, Rapaport DH, Rakic P. Cytogenesis in the monkeyretina. J Comp Neurol 1991; 309: 86–114.
7. Abramov I, Gordon J, Hendrickson A, Hainline L, Dobson V,LaBossiere E. The retina of the newborn human infant. Science1982; 217: 265–7.
8. Yuodelis C, Hendrickson A. A qualitative and quantitative analysisof the human fovea during development. Vision Res 1986; 26:847–55.
9. Hendrickson A, Drucker D. The development of parafoveal andmid-peripheral human retina. Behav Brain Res 1992; 49: 21–31.
10. Sjöstrand J, Popovic Z, Conradi N, Marshall J. Morphometricstudy of the displacement of retinal ganglion cells subserving coneswithin the human fovea. Graefes Arch Clin Exp Ophthalmol 1999;237: 1014–23.
11. Hendrickson AE, Yuodelis C. The morphological development ofthe human fovea. Ophthalmology 1984; 91: 603–12.
12. Carter-Dawson LD, LaVail MM. Rods and cones in the mouseretina. II. Autoradiographic analysis of cell generation usingtritiated thymidine. J Comp Neurol 1979; 188: 263–72.
13. Young RW. Cell differentiation in the retina of the mouse. Anat Rec1985; 212: 199–205.
14. Stiemke MM, Hollyfield JG. Cell birthdays in Xenopus laevisretina. Differentiation 1995; 58: 189–93.
15. Livesey FJ, Cepko CL. Vertebrate neural cell-fate determination:lessons from the retina. Nat Rev Neurosci 2001; 2: 109–18.
16. Belliveau MJ, Cepko CL. Extrinsic and intrinsic factors control thegenesis of amacrine and cone cells in the rat retina. Development1999; 126: 555–66.
17. Belliveau MJ, Young TL, Cepko CL. Late retinal progenitor cellsshow intrinsic limitations in the production of cell types and thekinetics of opsin synthesis. J Neurosci 2000; 20: 2247–54.
18. Cepko CL, Austin CP, Yang X, Alexiades M, Ezzeddine D. Cell fatedetermination in the vertebrate retina. Proc Natl Acad Sci USA1996; 93: 589–95.
19. Marquardt T, Gruss P. Generating neuronal diversity in the retina:one for nearly all. Trends Neurosci 2002; 25: 32–8.
20. Tomasiewicz H, Ono K, Yee D, Thompson C, Goridis C,Rutishauser U, et al. Genetic deletion of a neural cell adhesionmolecule variant (N-CAM-180) produces distinct defects in thecentral nervous system. Neuron 1993; 11: 1163–74.
21. Steinberg MS, Takeichi M. Experimental specification of cellsorting, tissue spreading, and specific spatial patterning byquantitative differences in cadherin expression. Proc Natl Acad SciUSA 1994; 91: 206–9.
22. Georges-Labouesse E, Mark M, Messaddeq N, Gansmuller A.Essential role of alpha 6 integrins in cortical and retinal lamination.Curr Biol 1998; 8: 983–6.
23. Scheibe R, Schnitzer J, Rohrenbeck J, Wohlrab F, Reichenbach A.Development of A-type (axonless) horizontal cells in the rabbitretina. J Comp Neurol 1995; 354: 438–58.
ch.02.qxd 29/3/04 05:49 AM Page 18
24. Reese BE, Necessary BD, Tam PP, Faulkner-Jones B, Tan SS. Clonalexpansion and cell dispersion in the developing mouse retina. Eur JNeurosci 1999; 11: 2965–78.
25. Galli-Resta L. Putting neurons in the right places: local interactionsin the genesis of retinal architecture. Trends Neurosci 2002; 25:638–43.
26. Jeffery G. Architecture of the optic chiasm and the mechanismsthat sculpt its development. Physiol Rev 2001; 81: 1393–414.
27. Stoeckli ET, Landmesser LT. Axon guidance at choice points. CurrOpin Neurobiol 1998; 8: 73–9.
28. Torres M, Gomez-Pardo E, Gruss P. Pax2 contributes to inner earpatterning and optic nerve trajectory. Development 1996; 122:3381–91.
29. Trousse F, Marti E, Gruss P, Torres M, Bovolenta P. Control ofretinal ganglion cell axon growth : a new role for Sonic hedgehog.Development 2001; 128: 3927–36.
30. Sretavan DW, Reichardt LF. Time-lapse video analysis of retinalganglion cell axon pathfinding at the mammalian optic chiasm:growth cone guidance using intrinsic chiasm cues. Neuron 1993;10: 761–77.
31. Sretavan DW, Pure E, Siegel MW, Reichardt LF. Disruption ofretinal axon ingrowth by ablation of embryonic mouse optic chiasmneurons. Science 1995; 269: 98–101.
32. Oster SF, Sretavan DW. Connecting the eye to the brain: themolecular basis of ganglion cell axon guidance. Br J Ophthalmol2003; 87: 639–45.
33. Plump AS, Erskine L, Sabatier C, Brose K, Epstein CJ, GoodmanCS, et al. Slit1 and Slit2 cooperate to prevent premature midlinecrossing of retinal axons in the mouse visual system. Neuron 2002;33: 219–32.
34. Herrera E, Brown L, Aruga J, Rachel RA, Dolen G, Mikoshiba K,et al. Zic2 Patterns binocular vision by specifying the uncrossedretinal projection. Cell 2003; 114: 545–57.
35. Fukuda Y, Sawai H, Watanabe M, Wakakuwa K, Morigiwa K.Nasotemporal overlap of crossed and uncrossed retinal ganglioncell projections in the Japanese monkey (Macaca fuscata). J Neurosci 1989; 9: 2353–73.
36. Guillery RW. Visual pathways in albinos. Sci Am 1974;230:44–54.37. Sincich LC, Horton JC. An albino-like decussation error in the
43. Hendry SH, Reid RC. The koniocellular pathway in primate vision.Annu Rev Neurosci 2000; 23: 127–53.
44. Rakic P. Prenatal genesis of connections subserving oculardominance in the rhesus monkey. Nature 1976; 261: 467–71.
45. Rakic P, Riley KP. Overproduction and elimination of retinal axonsin the fetal rhesus monkey. Science 1983; 219: 1441–4.
46. Provis JM. Patterns of cell death in the ganglion cell layer of thehuman fetal retina. J Comp Neurol 1987; 259: 237–46.
47. Sretavan DW, Shatz CJ. Prenatal development of individualretinogeniculate axons during the period of segregation. Nature1984; 308: 845–8.
48. Sretavan DW, Shatz CJ. Prenatal development of retinal ganglioncell axons: segregation into eye-specific layers within the cat’slateral geniculate nucleus. J Neurosci 1986; 6: 234–51.
49. Snider CJ, Dehay C, Berland M, Kennedy H, Chalupa LM.Prenatal development of retinogeniculate axons in the macaquemonkey during segregation of binocular inputs. J Neurosci 1998;19: 220–8.
50. Erkman L, McEvilly RJ, Luo L, Ryan AK, Hooshmand F, O’ConnellSM, et al. Role of transcription factors Brn-3.1 and Brn-3.2 inauditory and visual system development. Nature 1996; 381: 603–6.
51. Meissirel C, Wikler KC, Chalupa LM, Rakic P. Early divergence ofmagnocellular and parvocellular functional subsystems in theembryonic primate visual system. Proc Natl Acad Sci USA 1997;94: 5900–5.
54. Wiesel TN, Hubel DH, Lam DM. Autoradiographic demonstrationof ocular-dominance columns in the monkey striate cortex bymeans of transneuronal transport. Brain Res 1974; 79: 273–9.
55. Horton JC, Hocking DR. An adult-like pattern of oculardominance columns in striate cortex of newborn monkeys prior tovisual experience. J Neurosci 1996; 16: 1791–807.
56. LeVay S, Stryker MP, Shatz CJ. Ocular dominance columns andtheir development in layer IV of the cat’s visual cortex: aquantitative study. J Comp Neurol 1978; 179: 223–44.
58. Katz LC, Shatz CJ. Synaptic activity and the construction of corticalcircuits. Science 1996; 274: 1133–8.
59. Wong RO, Meister M, Shatz CJ. Transient period of correlatedbursting activity during development of the mammalian retina.Neuron 1993; 11: 923–38.
60. Stellwagen D, Shatz CJ. An instructive role for retinal waves in thedevelopment of retinogeniculate connectivity. Neuron 2002; 33:357–67.
61. Crowley JC, Katz LC. Development of ocular dominance columnsin the absence of retinal input. Nat Neurosci 1999; 2: 1125–30.
62. Sengpiel F, Kind PC. The role of activity in development of thevisual system. Curr Biol 2002; 12: R818–26.
63. van Essen DC. Organization of visual areas in macaque and humancerebral cortex. In: Werner JS, Chalupa LM, editors. Visualneurosciences. Cambridge, MA: MIT Press; 2003: 507–21.
64. Zeki S. A vision of the brain. Oxford: Blackwell, 1993.65. Huffman KJ, Molnar Z, van Dellen A, Kahn DM, Blakemore C,
Krubitzer L. Formation of cortical fields on a reduced corticalsheet. J Neurosci 1999; 19: 9939–52.
66. Valenza E, Simion F, Cassia VM, Umilta C. Face preference atbirth. J Exp Psychol Hum Percept Perform 1996; 22: 892–903.
67. Turati C, Simion F, Milani I, Umilta C. Newborns’ preference forfaces: what is crucial? Dev Psychol 2002; 38: 875–82.
69. Dobson V, Teller DY. Visual acuity in human infants: a review andcomparison of behavioral and electrophysiological studies. VisionRes 1978; 18: 1469–83.
70. Mayer DL, Dobson V. Visual acuity development in infants andyoung children, as assessed by operant preferential looking. VisionRes 1982; 22: 1141–51.
71. Berlyne DE. The influence of the albedo and complexity of stimulion visual fixation in the human infant. Br J Psychol 1958; 49:315–8.
72. Fantz RL. Pattern vision in young infants. Psychological Record1958; 8: 43–7.
73. Fantz RL, Ordy JM, Udelf MS. Maturation of pattern vision ininfants during the first six months. J Comp Physiol Psychol 1962;55: 907–17.
74. Teller DY, Morse R, Borton R, Regal D. Visual acuity for vertical anddiagonal gratings in human infants. Vision Res 1974; 14: 1433–9.
75. Atkinson J, Braddick O, Braddick F. Acuity and contrast sensitivityof infant vision. Nature 1974; 247: 403–4.
76. Banks MS, Salapatek P. Acuity and contrast sensitivity in 1-, 2-, and3-month-old human infants. Invest Ophthalmol Vis Sci 1978; 17:361–5.
77. Atkinson J, Braddick O, Pimm-Smith E. “Preferential looking” formonocular and binocular acuity testing of infants. Br J Ophthalmol1982; 66: 264–8.
78. Tyler CW, Apkarian P, Levi DM, Nakayama K. Rapid assessment ofvisual function: an electronic sweep technique for the patternvisual evoked potential. Invest Ophthalmol Vis Sci 1979; 18:703–13.
79. Norcia AM, Tyler CW. Spatial frequency sweep VEP: visual acuityduring the first year of life. Vision Res 1985; 25: 1399–408.
Normal and Abnormal Visual Development
CHAPTER
2
19
ch.02.qxd 29/3/04 05:49 AM Page 19
80. Harter MR, Suitt CD. Visually-evoked cortical responses andpattern vision in the infant: a longitudinal study. PsychonomicScience 1970; 18: 235–7.
81. Regan D. Evoked potentials in psychology, sensory physiology andclinical medicine. New York: Wiley-Interscience; 1972.
82. Harter MR, White CT. Evoked cortical responses to checkerboardpatterns: effect of check-size as a function of visual acuity.Electroencephalogr Clin Neurophysiol 1970; 28: 48–54.
83. Marg E, Freeman DN, Peltzman P, Goldstein PJ. Visual acuitydevelopment in human infants: evoked potential measurements.Invest Ophthalmol 1976; 15: 150–3.
84. Harter MR, Deaton FK, Odom JV. Maturation of evoked potentialsand visual preference in 6-45-day-old infants: effects of check size,visual acuity, and refractive error. Electroencephalogr ClinNeurophysiol 1977; 42: 595–607.
85. Harris L, Atkinson J, Braddick O. Visual contrast sensitivity of a 6-month-old infant measured by the evoked potential. Nature 1976;264: 570–1.
87. Gorman J, Cogan DG, Gellis SS. An apparatus for grading thevisual acuity on the basis of optokinetic nystagmus. Pediatrics1957; 19: 1088–92.
88. Dayton GO, Jones MH, Aiu P, Rawson RA, Steele B, Rose M.Developmental study of coordinated eye movements in the humaninfant. I. Visual activity in the newborn infant: a study based oninduced optokinetic nystagmus recorded by electro-oculography.Arch Ophthalmol 1964; 71: 865–70.
89. Riggs LA. Visual acuity. In: Graham CH, editor. Vision and visualperception. New York: Wiley; 1965: p. 321–49.
90. Gorman JJ, Cogan DG, Gellis SS. A device for testing visual acuityin infants. Sight Sav Rev 1959; 29: 80–4.
91. Miranda SB. Visual abilities and pattern preferences of prematureinfants and full-term neonates. J Exp Child Psychol 1970; 10:189–205.
92. McDonald MA, Dobson V, Sebris SL, Baitch L, Varner D, TellerDY. The acuity card procedure: a rapid test of infant acuity. InvestOphthalmol Vis Sci 1985; 26: 1158–62.
93. Teller DY, McDonald MA, Preston K, Sebris SL, Dobson V.Assessment of visual acuity in infants and children: the acuity cardprocedure. Dev Med Child Neurol 1986; 28: 779–89.
94. Courage ML, Adams RJ. Visual acuity assessment from birth tothree years using the acuity card procedure: cross-sectional andlongitudinal samples. Optom Vis Sci 1990; 67: 713–8.
95. Sokol S. Measurement of infant visual acuity from pattern reversalevoked potentials. Vision Res 1978; 18: 33–9.
96. Sokol S, Moskowitz A. Comparison of pattern VEPs andpreferential-looking behavior in 3-month-old infants. InvestOphthalmol Vis Sci 1985; 26: 359–65.
97. Schade OH. Optical and photoelectric analog of the eye. J Opt SocAm 1956; 46: 721–39.
98. Campbell FW, Robson JG. Application of Fourier analysis to thevisibility of gratings. J Physiol 1968; 197: 551–66.
99. Atkinson J, Braddick O, Moar K. Development of contrastsensitivity over the first 3 months of life in the human infant.Vision Res 1977; 17: 1037–44.
100. Kelly DH. Spatial frequency selectivity in the retina. Vision Res1975; 15: 665–72.
101. Eschweiler GW, Rauschecker JP. Temporal integration in visualcortex of cats with surgically induced strabismus. Eur J Neurosci1993; 5: 1501–9.
102. Hoekstra J, van der Goot DP, van den Brink G, Bilsen FA. Theinfluence of the number of cycles upon the visual contrast thresholdfor spatial sine wave patterns. Vision Res 1974; 14: 365–8.
103. Savoy RL, McCann JJ. Visibility of low-spatial-frequency sine-wavetargets: Dependence on number of cycles. J Opt Soc Am 1975; 65:343–50.
104. Banks MS, Salapatek P. Contrast sensitivity function of the infantvisual system. Vision Res 1976; 16: 867–9.
105. Norcia AM, Tyler CW, Allen D. Electrophysiological assessment ofcontrast sensitivity in human infants. Am J Optom Physiol Opt1986; 63: 12–5.
106. Norcia AM, Tyler CW, Hamer RD. High visual contrast sensitivityin the young human infant. Invest Ophthalmol Vis Sci 1988; 29:44–9.
107. Norcia AM, Tyler CW, Hamer RD. Development of contrastsensitivity in the human infant. Vision Res 1990; 30: 1475–86.
108. Campbell FW, Maffei L. Electrophysiological evidence for theexistence of orientation and size detectors in the human visualsystem. J Physiol 1970; 207: 635–52.
109. Teller DY, Regal DM, Videen TO, Pulos PE. Development of visualacuity in infant monkeys (Macaca nemestrina) during the earlypostnatal weeks. Vision Res 1978; 18: 561–6.
110. Boothe RG, Williams RA, Kiorpes L, Teller DY. Development ofcontrast sensitivity in infant Macaca nemestrina monkeys. Science1980; 208: 1290–2.
111. Kiorpes L, Kiper DC. Development of contrast sensitivity acrossthe visual field in macaque monkeys (Macaca nemestrina). VisionRes 1996; 36: 239–47.
112. Allen D, Tyler CW, Norcia AM. Development of grating acuity andcontrast sensitivity in the central and peripheral visual field of thehuman infant. Vision Res 1996; 36: 1945–53.
113. Kiorpes L, Movshon JA. Peripheral and central factors limiting thedevelopment of contrast sensitivity in macaque monkeys. VisionRes 1998; 38: 61–70.
114. Salapatek P, Bechtold AG, Bushnell EW. Infant visual acuity as afunction of viewing distance. Child Dev 1976; 47: 860–3.
115. Julesz B. Foundations of cyclopean perception. Chicago: Univ. ofChicago Press; 1971.
116. Tomac S, Birdal E. Effects of anisometropia on binocularity. JPediatr Ophthalmol Strabismus 2001; 38: 27–33.
117. Julesz B. Binocular depth perception of computer generatedpatterns. Bell Systems Technical Journal 1960; 39: 1125–62.
118. Atkinson J, Braddick O. Stereoscopic discrimination in infants.Perception 1976; 5: 29–38.
119. Fox R, Aslin RN, Shea SL, Dumais ST. Stereopsis in human infants.Science 1980; 207: 323–4.
120. Shea SL, Fox R, Aslin RN, Dumais ST. Assessment of stereopsis inhuman infants. Invest Ophth Vis Sci 1980; 19: 1400–4.
121. Julesz B, Kropfl W, Petrig B. Large evoked potentials to dynamicrandom-dot correlograms and stereograms permit quick deter-mination of stereopsis. Proc Natl Acad Sci USA 1980; 77: 2348–51.
122. Braddick O, Atkinson J, Julesz B, Kropfl W, Bodis-Wollner I, RaabE. Cortical binocularity in infants. Nature 1980; 288: 363–5.
123. Petrig B, Julesz B, Kropfl W, Baumgartner G, Anliker M.Development of stereopsis and cortical binocularity in humaninfants: Electrophysiological evidence. Science 1981; 213: 1402–5.
124. Braddick O, Wattam-Bell J, Day J, Atkinson J. The onset ofbinocular function in human infants. Hum Neurobiol 1983; 2: 65–9.
125. Birch EE. Stereopsis in children and its developmental relation tovisual acuity. In: Simons K, editor. Early visual development,normal and abnormal. New York: Oxford University Press; 1993:224–36.
126. Birch E, Petrig B. FPL and VEP measures of fusion, stereopsis andstereoacuity in normal infants. Vision Res 1996; 36: 1321–7.
127. Ciner EB, Schanel-Klitsch E, Herzberg C. Stereoacuity devel-opment: 6 months to 5 years. A new tool for testing and screening.Optom Vis Sci 1996; 73: 43–8.
128. Pettigrew JD, Nikara T, Bishop PO. Binocular interaction on singleunits in cat striate cortex: simultaneous stimulation by singlemoving slit with receptive fields in correspondence. Exp Brain Res1968; 6: 391–410.
129. Poggio GF, Fischer B. Binocular interaction and depth sensitivity instriate and prestriate cortex of behaving rhesus monkey. JNeurophysiol 1977; 40: 1392–405.
130. Cumming BG, Parker AJ. Responses of primary visual corticalneurons to binocular disparity without depth perception. Nature1997; 389: 280–3.
131. Cumming BG, Parker AJ. Binocular neurons in V1 of awakemonkeys are selective for absolute, not relative, disparity. JNeurosci 1999; 19: 5602–18.
132. Chino YM, Smith EL 3rd, Hatta S, Cheng H. Postnatal developmentof binocular disparity sensitivity in neurons of the primate visualcortex. J Neurosci 1997; 17: 296–307.
133. O’Dell C, Boothe RG. The development of stereoacuity in infantrhesus monkeys. Vision Res 1997; 37: 2675–84.
134. Hubel DH, Wiesel TN. Receptive fields of single neurones in thecat’s striate cortex. J Physiol (London) 1959; 148: 574–91.
135. Barlow HB, Pettigrew JD. Lack of specificity of neurones in thevisual cortex of young kittens. J Physiol 1971;218:98–101.
EPIDEMIOLOGY, GROWTH AND DEVELOPMENT
SECTION
1
20
ch.02.qxd 29/3/04 05:49 AM Page 20
136. Wiesel TN, Hubel DH. Ordered arrangement of orientationcolumns in monkeys lacking visual experience. J Comp Neurol1974; 158: 307–18.
137. Blasdel G, Obermayer K, Kiorpes L. Organization of oculardominance and orientation columns in the striate cortex ofneonatal macaque monkeys. Vis Neurosci 1995; 12: 589–603.
138. Blakemore C, Cooper GF. Development of the brain depends onthe visual environment. Nature 1970; 228: 477–8.
139. Hirsch HV. Visual perception in cats after environmental surgery.Exp Brain Res 1972; 15: 405–23.
140. Blasdel GG, Mitchell DE, Muir DW, Pettigrew JD. A physiologicaland behavioural study in cats of the effect of early visualexperience with contours of a single orientation. J Physiol(London) 1977; 265: 615–36.
141. Stryker MP, Sherk H. Modification of cortical orientationselectivity in the cat by restricted visual experience: a re-examination. Science 1975; 190: 904–6.
142. Stryker MP, Sherk H, Leventhal AG, Hirsch HV. Physiologicalconsequences for the cat’s visual cortex of effectively restrictingearly visual experience with oriented contours. J Neurophysiol1978; 41: 896–909.
143. Nakayama K. Biological image motion processing: a review. VisionRes 1985; 25: 625–60.
144. Kremenitzer JP, Vaughan HG Jr, Kurtzberg D, Dowling K. Smooth-pursuit eye movements in the newborn infant. Child Dev 1979;50: 442–8.
145. Wattam-Bell J. Visual motion processing in one-month-old infants: preferential looking experiments. Vision Res 1996; 36:1671–7.
146. Wattam-Bell J. Visual motion processing in one-month-old infants:habituation experiments. Vision Res 1996; 36: 1679–85.
147. Wattam-Bell J. Development of motion-specific cortical responsesin infancy. Vision Res 1991; 31: 287–97.
148. Hamer RD, Norcia AM. The development of motion sensitivityduring the first year of life. Vision Res 1994; 34: 2387–402.
149. Banton T, Dobkins K, Bertenthal BI. Infant direction discriminationthresholds. Vision Res 2001; 41: 1049–56.
150. Hawken MJ, Parker AJ, Lund JS. Laminar organization andcontrast selectivity of direction selective cells in the striate cortexof the Old-World monkey. J Neurosci 1988; 8: 3541–8.
151. Zeki SM. Functional organization of a visual area in the posteriorbank of the superior temporal sulcus of the rhesus monkey. J Physiol (London) 1974; 236: 549–73.
153. Hatta S, Kumagami T, Qian J, Thornton M, Smith EL 3rd, ChinoYM. Nasotemporal directional bias of V1 neurons in young infantmonkeys. Invest Ophthalmol Vis Sci 1998; 39: 2259–67.
275–7.157. Bornstein MH. Qualities of color vision in infancy. J Exp Child
Psychol 1975; 19: 401–19.158. Peeples DR, Teller DY. Color vision and brightness discrimination
in two-month-old human infants. Science 1975; 189: 1102–3.159. Brown AM. Development of visual sensitivity to light and color vision
in human infants: a critical review. Vision Res 1990; 30: 1159–88.160. Teller DY, Peeples DR, Sekel M. Discrimination of chromatic from
white light by two-month-old human infants. Vision Res 1978; 18:41–8.
161. Hamer RD, Alexander KR, Teller DY. Rayleigh discriminations inyoung human infants. Vision Res 1982; 22: 575–7.
162. Packer O, Hartmann EE, Teller DY. Infant color vision: the effectof test field size on Rayleigh discriminations. Vision Res 1984; 24:1247–60.
163. Boynton RM. Human color vision, special limited. Washington,DC: Optical Society of America; 1992.
164. Banks MS, Bennett PJ. Optical and photoreceptor immaturitieslimit the spatial and chromatic vision of human neonates. J OptSoc Am A 1988; 5: 2059–79.
165. Morrone MC, Speed HD, Burr DC. Development of visualinhibitory interactions in kittens. Vis Neurosci 1991; 7: 321–34.
166. Movshon JA, Kiorpes L. Analysis of the development of spatial
contrast sensitivity in monkey and human infants. J Opt Soc Am A1988; 5: 2166–72.
167. Morrone MC, Burr DC, Fiorentini A. Development of infantcontrast sensitivity to chromatic stimuli. Vision Res 1993; 33:2535–52.
168. Morrone MC, Fiorentini A, Burr DC. Development of thetemporal properties of visual evoked potentials to luminance andcolour contrast in infants. Vision Res 1996; 36: 3141–55.
169. Kelly JP, Borchert K, Teller DY. The development of chromatic andachromatic contrast sensitivity in infancy as tested with the sweepVEP. Vision Res 1997; 37: 2057–72.
170. Wiesel TN, Hubel DH. Single cell responses in striate cortex ofkittens deprived of vision in one eye. J Neurophysiol 1963; 26:1003–17.
171. Hubel DH, Wiesel TN. The period of susceptibility to thephysiological effects of unilateral eye closure in kittens. J Physiol(London) 1970; 206: 419–36.
172. Daw NW, Fox K, Sato H, Czepita D. Critical period for monoculardeprivation in the cat visual cortex. J Neurophysiol 1992; 67:197–202.
173. Levitt FB, van Sluyters RC. The sensitive period for strabismus inthe kitten. Brain Res 1982; 255: 323–7.
174. Daw NW, Wyatt HJ. Kittens reared in a unidirectional environment:evidence for a critical period. J Physiol 1976; 257: 155–70.
175. Kim DS, Bonhoeffer T. Reverse occlusion leads to a preciserestoration of orientation preference maps in visual cortex. Nature1994; 370: 370–2. (Published erratum appears in Nature 1994Nov 10;372(6502): 196.)
176. Shatz CJ, Stryker MP. Ocular dominance in layer IV of the cat’svisual cortex and the effects of monocular deprivation. J Physiol(London) 1978; 281: 267–83.
177. Harwerth RS, Smith EL, 3rd, Duncan GC, et al. Multiple sensitiveperiods in the development of the primate visual system. Science1986; 232: 235–8.
178. Vaegan, Taylor D. Critical period for deprivation amblyopia inchildren. Trans Ophthalmol Soc UK 1979; 99: 432–9.
179. Vereecken EP, Brabant P. Prognosis for vision in amblyopia after theloss of the good eye. Arch Ophthalmol 1984; 102: 220–4.
180. Tierney DW. Vision recovery in amblyopia after contralateralsubretinal hemorrhage. J Am Optom Assoc 1989; 60: 281–3.
181. Rahi JS, Logan S, Borja MC, Timms C, Russell-Eggitt I, Taylor D.Prediction of improved vision in the amblyopic eye after visual lossin the non-amblyopic eye. Lancet 2002; 360: 621–2.
182. Lepard CW. Comparative changes in the error of refraction betweenfixing and amblyopic eyes during growth and development. Am JOphthalmol 1975; 80: 485–90.
183. von Noorden GK, Campos EC. Binocular vision and ocularmotility: theory and management of strabismus. 6th ed. St. Louis,MO: Mosby; 2002.
184. Archer SM, Sondhi N, Helveston EM. Strabismus in infancy.Ophthalmology 1989; 96: 133–7.
185. Sondhi N, Archer SM, Helveston EM. Development of normalocular alignment. J Pediatr Ophthalmol Strabismus 1988; 25: 210–1.
186. Gwiazda J, Thorn F. Development of refraction and strabismus.Curr Opin Ophthalmol 1999; 10: 293–9.
189. Abrahamsson M, Fabian G, Sjostrand J. A longitudinal study of apopulation based sample of astigmatic children. II. Thechangeability of anisometropia. Acta Ophthalmol (Copenhagen)1990; 68: 435–40.
190. Boothe RG, Teller DY. Meridional variations in acuity and CSFs inmonkeys (Macaca nemestrina) reared with externally appliedastigmatism. Vision Res 1982; 22: 801–10.
191. Hendrickson AE, Movshon JA, Eggers HM, Gizzi MS, Boothe RG,Kiorpes L, et al. Effects of early unilateral blur on the macaque’svisual system. II. Anatomical observations. J Neurosci 1987; 7:1327–39.
192. Kiorpes L, Boothe RG, Hendrickson AE, Movshon JA, Eggers HM,Gizzi MS. Effects of early unilateral blur on the Macaque’s visual system. 1. Behavioral observations. J Neurosci 1987; 7:1318–26.
Normal and Abnormal Visual Development
CHAPTER
2
21
ch.02.qxd 29/3/04 05:49 AM Page 21
193. Movshon JA, Eggers HM, Gizzi MS, Hendrickson AE, Kiorpes L,Boothe RG. Effects of early unilateral blur on the macaque’s visualsystem. III. Physiological observations. J Neurosci 1987; 7: 1340–51.
194. Atkinson J, Braddick O, French J. Infant astigmatism: itsdisappearance with age. Vision Res 1980; 20: 891–3.
195. Gwiazda J, Scheiman M, Mohindra I, Held R. Astigmatism inchildren: changes in axis and amount from birth to six years. InvestOphthalmol Vis Sci 1984; 25: 88–92.
196. Saunders KJ. Early refractive development in humans. SurvOphthalmol 1995; 40: 207–16.
197. Ehrlich DL, Braddick OJ, Atkinson J, Anker S, Weeks F, Hartley T,et al. Infant emmetropization: longitudinal changes in refractioncomponents from nine to twenty months of age. Optom Vis Sci1997; 74: 822–43.
198. Dobson V, Miller JM, Harvey EM, Sherrill DL. Amblyopia inastigmatic preschool children. Vision Res 2003; 43: 1081–90.
199. Freeman RD, Mitchell DE, Millodot M. A neural effect of partialvisual deprivation in humans. Science 1972; 175: 1384–6.
200. Mitchell DE, Freeman RD, Millodot M, Haegerstrom G.Meridional amblyopia: evidence for modification of the humanvisual system by early visual experience. Vis Res 1973; 13: 535–58.
201. Harweth RS, Smith EL 3rd, Boltz RL. Meridional amblyopia inmonkeys. Exp Brain Res 1980; 39: 351–6.
202. Jacobson SG, Mohindra I, Held R. Development of visual acuity ininfants with congenital cataracts. Br J Ophthalmol 1981; 65: 727–35.
203. Birch EE, Stager DR, Wright WW. Grating acuity developmentafter early surgery for congenital unilateral cataract. ArchOphthalmol 1986; 104: 1783–7.
204. Drummond GT, Scott WE, Keech RV. Management of monocularcongenital cataracts. Arch Ophthalmol 1989; 107: 45–51.
205. Taylor D, Wright KW, Amaya L, Cassidy L, Nischal K, Russell-EggittI, et al. Should we aggressively treat unilateral congenital cataracts?Br J Ophthalmol 2001; 85: 1120–6.
206. Smith EL 3rd, Hung LF, Harwerth RS. The degree of imagedegradation and the depth of amblyopia. Invest Ophthalmol VisSci 2000; 41: 3775–81.
207. von Noorden GK. Classification of amblyopia. Am J Ophthalmol1967; 63: 238–44.
208. Schapero M. Amblyopia. 1st ed. Philadelphia: Chilton; 1971.209. Bradley A, Freeman RD. Is reduced Vernier acuity in amblyopia due
to position, contrast or fixation deficits? Vision Res 1985; 25: 55–66.210. Levi DM, Klein SA. Vernier acuity, crowding and amblyopia. Vision
Res 1985; 25: 979–91.211. Hess RF, Holliday IE. The spatial localization deficit in amblyopia.
Vision Res 1992; 32: 1319–39.212. Birch EE, Swanson WH. Hyperacuity deficits in anisometropic and
strabismic amblyopes with known ages of onset. Vision Res 2000;40: 1035–40.
213. McKee SP, Levi DM, Movshon JA. The pattern of visual deficits inamblyopia. J Vision 2003; 3: 380–405.
214. Levi DM, Klein S. Hyperacuity and amblyopia. Nature 1982; 298:268–70.
215. Levi DM, Klein S. Differences in Vernier discrimination for gratingbetween strabismic and anisometropic amblyopes. InvestOphthalmol Vis Sci 1982; 23: 398–407.
216. Levi DM, Klein SA, Yap YL. Positional uncertainty in peripheraland amblyopic vision. Vision Res 1987; 27: 581–97.
217. Levi DM, Klein SA, Wang H. Amblyopic and peripheral Vernier acuity: a test-pedestal approach. Vision Res 1994; 34:3265–92.
218. Levi DM, Klein SA, Wang H. Discrimination of position and contrastin amblyopic and peripheral vision. Vision Res 1994; 34: 3293–313.
219. Westheimer G, Hauske G. Temporal and spatial interference withVernier acuity. Vision Res 1975; 15: 1137–41.
220. Butler TW, Westheimer G. Interference with stereoscopic acuity:Spatial, temporal, and disparity tuning. Vision Res 1978; 18:1387–92.
221. Flom MC, Weymouth FW, Kahnemann D. Visual resolution andcontour integration. J Ophthalmol Soc Am 1963; 53: 1026–32.
222. Hess RF, Jacobs RJ. A preliminary report of acuity and contourinteractions across the amblyope’s visual field. Vision Res 1979; 19:1403–8.
223. Flom MC, Bedell HE, Allen J. Contour integration and visualresolution: Contralateral effets. Science 1963; 142: 979–80.
224. Barlow HB. The Ferrier Lecture, 1980. Critical limiting factors inthe design of the eye and visual cortex. Proc R Soc Lond B Biol Sci1981; 212: 1–34.
225. Westheimer G. Visual hyperacuity. Prog Sensory Physiol 1981; 1:1–30.
226. Wiesel TN. Postnatal development of the visual cortex and theinfluence of environment. Nature 1982; 299: 583–91.
227. Smith EL 3rd, Chino YM, Ni J, Cheng H, Crawford ML, HarwerthRS. Residual binocular interactions in the striate cortex of monkeysreared with abnormal binocular vision. J Neurophysiol 1997; 78:1353–62.
228. Kiorpes L, Kiper DC, O’Keefe LP, Cavanaugh JR, Movshon JA.Neuronal correlates of amblyopia in the visual cortex of macaquemonkeys with experimental strabismus and anisometropia. J Neurosci 1998; 18: 6411–24.
229. Sharma V, Levi DM, Klein SA. Undercounting features and missingfeatures: evidence for a high-level deficit in strabismic amblyopia.Nat Neurosci 2000; 3: 496–501.
230. Wiesel TN, Hubel DH. Comparison of the effects of unilateral andbilateral eye closure on cortical unit responses in kittens. J Neurophysiol 1965; 28: 1029–40.
231. Wiesel TN, Hubel DH. Extent of recovery from the effects ofvisual deprivation in kittens. J Neurophysiol 1965; 28: 1060–72.
232. Hubel DH, Wiesel TN, LeVay S. Plasticity of ocular dominancecolumns in monkey striate cortex. Philos Trans R Soc Lond B BiolSci 1977; 278: 377–409.
233. LeVay S, Wiesel TN, Hubel DH. The development of oculardominance columns in normal and visually deprived monkeys. J Comp Neurol 1980; 191: 1–51.
234. Horton JC, Stryker MP. Amblyopia induced by anisometropiawithout shrinkage of ocular dominance columns in human striatecortex. Proc Natl Acad Sci USA 1993; 90: 5494–8.
235. Horton JC, Hocking DR. Pattern of ocular dominance columns inhuman striate cortex in strabismic amblyopia. Vis Neurosci 1996;13: 787–95.
236. Horton JC, Hocking DR, Kiorpes L. Pattern of ocular dominancecolumns and cytochrome oxidase activity in a macaque monkeywith naturally occurring anisometropic amblyopia. Vis Neurosci1997; 14: 681–9.
237. Goodyear BG, Nicolle DA, Menon RS. High resolution fMRI ofocular dominance columns within the visual cortex of humanamblyopes. Strabismus 2002; 10: 129–36.
238. Löwel S, Singer W. Experience dependant plasticity of intracorticalconnections. In: Fahle M, Poggio T, editors. Perceptual learning.Cambridge, MA: MIT Press; 2002: 3–18.
239. Löwel S, Engelmann R. Neuroanatomical and neurophysiologicalconsequences of strabismus: changes in the structural and functionalorganization of the primary visual cortex in cats with alternatingfixation and strabismic amblyopia. Strabismus 2002; 10:95–105.
240. Dobson V. Visual acuity testing by preferential looking techniques.In: Isenberg SJ, editor.The eye in infancy. St. Louis, MO: Mosby;1994. 131–56.
241. Gwiazda J, Brill S, Mohindra I, Held R. Infant visual acuity and itsmeridional variation. Vision Res 1978; 18: 1557–64.
242. Allen JL. The development of visual acuity in human infants duringthe early postnatal weeks. Seattle: University of Washington; 1979.
243. Van Hof-Van Duin J, Mohn G. Monocular and binocular optokineticnystagmus in humans with defective stereopsis. Invest OphthalmolVis Sci 1986; 27: 574–83.
244. Birch EE, Hale LA. Criteria for monocular acuity deficit in infancyand early childhood. Invest Ophthalmol Vis Sci 1988; 29: 636–43.
245. McDonald M, Ankrum C, Preston K, Sebris SL, Dobson V. Mono-cular and binocular acuity estimation in 18- to 36-month-olds: acuitycard results. Am J Optom Physiol Opt 1986; 63: 181–6.
